Why do we have two archive pages, Wikipedia talk:Automated taxobox system/Archive 1 and Wikipedia talk:Automated taxobox system/Archive 13, and neither link to each other and only one is linked above? --Nessie (talk) 19:19, 27 February 2019 (UTC)

@Peter coxhead:, looks like you set it up. --Nessie (talk) 18:29, 6 March 2019 (UTC)

Clearly the behaviour of the bot is beyond my understanding; I hope someone can sort it out! Peter coxhead (talk) 18:33, 6 March 2019 (UTC)

@Peter coxhead: As NessieVL explained above, you told the bot so start at archive 13 using |numberstart=13 in the archive template. I've fixed the problem. {{3x|p}}ery (talk) 21:17, 9 March 2019 (UTC)

@Pppery: how exceedingly silly of me! Thanks for sorting this out. Peter coxhead (talk) 21:23, 9 March 2019 (UTC)

See Wikipedia talk:Template namespace#RfC on templates storing data. The automated taxobox system currently relies on storing the taxonomic hierarchy in taxonomy templates, which one outcome of this RfC would forbid. Peter coxhead (talk) 10:11, 16 March 2019 (UTC)

Please remove the line <noinclude>{{pp-template|small=yes}}</noinclude> as well as [[Category:Taxobox templates]] - protection templates are automatically handled by the documentation page, and the documentation page already has the category. Thanks, --DannyS712 (talk) 06:55, 22 March 2019 (UTC)

 Done — JJMC89 (T·C) 07:02, 22 March 2019 (UTC)

Hi, I've just created Template:Taxonomy/Mini to go with Draft:Mini (genus) (or Draft:Mini (frog) if that ends up being a better idea). Have I done the right thing for the genus link in the taxobox? I can't put Mini in there. Newystats (talk) 20:21, 28 March 2019 (UTC)

I've changed to link to go to the draft page, as it makes sense for the draft to display the intended appearance. The link will need to be changed when the article is moved to main space.   Jts1882 | talk  20:45, 28 March 2019 (UTC)

Newystats, you should use a pipe so to hide the disambiguator. E.g., for the link line enter "Mini (genus)|Mini" (minus the quotes).

As for (genus) vs. (frog), most genera that require disambiguation are using a term such as (frog) rather than (genus). The single biggest source of ambiguity for animal genera is a plant/alga/fungus genus with the same name (see Category:Genus disambiguation pages), so (genus) can't be used very consistently. I'm pretty set against (genus) as a disambiguator, but I've come around to consider that (genus) might be a decent choice in the case of very short (2-4 letter) genus names. Longer genus names are probably generally recognizable as being the scientific name of something (even if you have no idea what), which makes (genus) redundant on top of it not always be sufficient to disambiguate against a genus in a different kingdom. Short genus names like Mini, Pan, and Sus aren't immediately recognizable as being scientific names. Plantdrew (talk) 21:10, 28 March 2019 (UTC)

I think it's better to be completely consistent and avoid "(genus)" in favour of a more informative disambiguator, such as "(frog)" in this case. Peter coxhead (talk) 22:29, 28 March 2019 (UTC)

Thanks All, Draft moved to (frog). Newystats (talk) 23:56, 28 March 2019 (UTC)

Like with Module:Documentation, this module should automatically sense if a page is protected, and transclude the appropriate icon and category. There are currently 176 template-protected taxobox templates, and likely many more with other levels of protection. I believe that implementing this should be non-controversial, but have no idea how it would be done, and don't want to try and guess on such a widely used module. Thanks, --DannyS712 (talk) 00:16, 29 March 2019 (UTC)

 Not done: it's not clear what changes you want to be made. Please mention the specific changes in a "change X to Y" format and provide a reliable source if appropriate. If you don't know how to do it, I would recommend asking the authors of the documentation module. -- /Alex/₂₁ 12:19, 29 March 2019 (UTC)

The speciesbox at Long-tailed mole is showing tribe. It should be showing tribe; no problem there. But I don't understand why it is showing; |display_parents= is lacking and the taxonomy template for the tribe doesn't have |always_display=. Did something change recently? If so, what is the new logic for displaying minor ranks? Plantdrew (talk) 20:23, 1 April 2019 (UTC)

@Plantdrew: it displays because |grandparent_authority= is set; if you remove this parameter, the tribe row disappears. It's always been the case, as far as I know, that in an automated taxobox every taxon with an authority given is displayed. Peter coxhead (talk) 09:06, 2 April 2019 (UTC)

It should display the subfamily, because of the monotypic tribe though. --Nessie (talk) 20:00, 2 April 2019 (UTC)

Converting {{automatic taxobox}} to Lua has caused it to display differently on some pages. If there is a {{DISPLAYTITLE:}} or {{italic title}} above it, there is extra vertical space at the top of the page above the taxobox where there wasn't before. This can be seen in Silphium (genus), for instance. This is because the MediaWiki software inserts a newline before the table-starting syntax ({|) if it appears at the beginning of the text that is output by a module. The display title is interpreted as an empty line by the parser, so there are now two empty lines, which generate an empty p tag above the table, which displays as blank vertical space.

This has also been an issue for {{Weather box}} (see Template talk:Weather box § Spacing). To prevent this, the table is enclosed in a div tag, with a newline before {| (see the main function in Module:Weather box). This both ensures that no extra whitespace is displayed and that the table syntax is interpreted correctly. — Eru·tuon 06:14, 2 April 2019 (UTC)

I think this is another case of the parse bug that we had with the {{clade}} module. Adding a <nowiki/> statement before the invoke statement was the workaround. A quick test shows this removes the line when added before the {{automatic taxobox}} in Silphium (genus).   Jts1882 | talk  08:18, 2 April 2019 (UTC)

 Done @Erutuon: I've added the <nowiki/> statement to {{automatic taxobox}} before the invoke statement. This removes the line in the example. The parser bug is discussed at T18700.   Jts1882 | talk  08:32, 2 April 2019 (UTC)

@Jts1882: thanks; this was, I suspect, a bug I re-introduced when converting the template to Lua. (Btw, DISPLAYTITLE isn't needed in cases like Silphium (genus); the automatically added {{Italic title}} handles simple disambiguation.) Peter coxhead (talk) 09:11, 2 April 2019 (UTC)

@Jts1882: Thanks! I had seen the parser bug mentioned in the source code of another template, but didn't connect it to what was happening here. — Eru·tuon 20:37, 2 April 2019 (UTC)

The documentation for {{Automatic taxobox}} is currently split between these two pages:

1. Template:Automatic taxobox/doc
2. Wikipedia:Automated taxobox system/new [Now merged into (1) as per below.]

(The odd title for the second page is because it was at "Template:Automatic taxobox/doc/new", but this page was never transcluded, only wikilinked, and some editors involved in TfD, etc. objected, saying that pages in template space have to have transclusions, so I reluctantly moved it.)

I've tidied up Wikipedia:Automated taxobox system/new, but my question is whether it's now useful to have two pages to document this template. When it was first set up, {{Automatic taxobox}} seems to have been the only automated taxobox template, so the documentation had two purposes: to document the template and introduce the automated taxobox system. Now there's plenty of documentation on the system generally, so I'm inclined to merge the two pages to Template:Automatic taxobox/doc, removing stuff that is common to other automated taxoboxes.

Views? Peter coxhead (talk) 19:46, 3 April 2019 (UTC)

• Difficult to see a reason not to merge, but a comment: the first thing that needs to be established is that a user is probably at the wrong page and they likely want Template:Speciesbox/doc instead (as I understand it). I would also prefer the string of exclamation marks be reduced from four to three. cygnis insignis 21:01, 3 April 2019 (UTC)
• Merge. Additionally, I would make it clear in the sidebar that the links there are to additional documentation and explanation. Right now, it looks like the "new" page has two TOCs. – Jonesey95 (talk) 04:03, 4 April 2019 (UTC)

@Jonesey95: in principle, it's already clear, from the title of the nav template/sidebar. However, I accept that for many readers it's not. Initially "automatic taxobox" was used to refer to the system as a whole, not just taxoboxes created by {{Automatic taxobox}} (because this was the only such template) – which is why all the documentation was at "Template:Automatic taxobox/doc" or its subpages. More recently, as {{Speciesbox}} became the most used template in the system, and as others were created, like {{Subspeciesbox}} and {{Virusbox}}, "automated taxobox" was used to refer to all the taxoboxes that rely on taxonomy templates, and "automated taxobox system" to the system as a whole. If you notice and accept these distinctions, the sidebar nav template's title does make it clear that it's not about {{Automatic taxobox}} but about the automated taxobox system. However, as I said, readers don't always seem to notice. Does anyone have any ideas as to how we can make the distinction clearer? Peter coxhead (talk) 08:25, 4 April 2019 (UTC)

Add a caption above the box. Something like "Further information", "Further information on the automated taxobox system", "Automated taxobox system documentation", or something pithier.   Jts1882 | talk  09:04, 4 April 2019 (UTC)

Ok, I've changed the box heading to "Documentation for the automated taxobox system". It's at Template:Autotaxobox system nav if anyone wants to try something different. Peter coxhead (talk) 09:19, 4 April 2019 (UTC)

How about putting it outside the box as it applies to the whole box, not just the top section (e.g. see {{Autotaxobox system navbox/sandbox}}).   Jts1882 | talk  10:12, 4 April 2019 (UTC)

I guess I saw it as like an article, where the lead section doesn't have its own heading, but I see what you mean. How about {{Autotaxobox system navbox/sandbox}} now? Peter coxhead (talk) 10:20, 4 April 2019 (UTC)

The lead section doesn't need a heading. I meant that the heading looked like it only applied to the top section rather than the whole box.   Jts1882 | talk  13:07, 4 April 2019 (UTC)

Ok, so as it is now: Template:Autotaxobox system navbox/sandbox? Peter coxhead (talk) 13:28, 4 April 2019 (UTC)

I think so.   Jts1882 | talk  13:56, 4 April 2019 (UTC)

As no-one dissented, I've merged the two documentation pages to Template:Automatic taxobox/doc – it's very much a first step, so please feel free to edit it. The material at Template:Automatic taxobox/doc#"Missing taxonomy template" applies to any automated taxobox; maybe it should be templated and repeated, particularly at {{Speciesbox/doc}}? Peter coxhead (talk) 13:56, 4 April 2019 (UTC)

I added a description to the template data table at Template:Automatic taxobox/doc in response to the recent discussion at Wikipedia talk:Manual of Style/Organisms#Wikipedia not following Code recommendations; I don't think we need to, but should document that we're not following. Peter coxhead (talk) 09:33, 4 April 2019 (UTC)

So I just beheld the wonder that is Vermetidae incertae sedis goreensis and tried to upgrade the box to a {{speciesbox}}. For some reason it doesn't want to make pretty the hot mess that Johann Friedrich Gmelin gave it for the 'binomial.' Any ideas? Oh, it's also a nomen dubium and a stub. --Nessie (talk) 02:59, 5 April 2019 (UTC)

As a nomen dubium, I would have nominated it for deletion as a non-notable subject (after all, it's not recognized as a species), but that failed the last time I tried it (q.v. this AfD for a taxon inquirendum) so I predict with that precedent, we are now stuck with potentially thousands of useless stubs for names that are not recognised as species. Perhaps the best that we can do with these is redirect them to a synonymous article page, but that has to exist first. Loopy30 (talk) 03:33, 5 April 2019 (UTC)

It isn't the name that Gmelin gave it; that was Serpula goreensis. IRMNG has it as Vermetus goreensis (Gmelin, 1791), which is the name of the Wikidata item. What WoRMS is saying is that it doesn't accept that the species is in the genus Vermetus, but doesn't know what genus it should be in. There are two possibilities, in my view:

• Move it to Vermetus goreensis, explaining in the article that the genus isn't accepted by WoRMS. In which case there can be a taxobox.
• Leave it as it is, but without a taxobox (as I've left it), because such a badly formed name isn't a taxon name and shouldn't be the target of a taxobox.

Peter coxhead (talk) 14:50, 5 April 2019 (UTC)

Vermetus is a real mess on WoRMs. Three other species are incertae sedis and 16 are taxa inquirenda. Move this one to Vermetus goreensis, which has actually been published, and explain the status in the article. Plantdrew (talk) 15:16, 5 April 2019 (UTC)

As several of you are aware I've been looking at ways of traversing the taxonomy templates downwards and generating a tree of the taxonomic hierarchy contained in the taxonomy templates. The Wikipedia API can be used to search child taxonomy templates of a taxon, i.e. taxonomy templates with |. The children can then be used to generate a listing of the hierarchy. Using this approach I have written a user script taxonomybrowser.js that adds a "taxonomy browser" option to the tool menu, which opens a dialog that can be used to browse the templates in the automatic taxonomy system. The script can be installed by adding the following line to your user common.js. For more on user scripts see Wikipedia:User_scripts.

 importScript('User:Jts1882/taxonomybrowser.js'); 

To get the tree for a taxon, open the taxonomy browser and enter the taxon. A default value can be got from the page in some circumstances, either from a taxonomy template page or from a mainspace article page named after the taxon. The tree can be navigated interactively using the ⊞ and ⊟ symbols. Various options allow hiding of extinct taxa or highlighting the taxa displayed in an automatic taxobox. A tooltip showing the parameters in the templates is available to help map the taxonomy templates and their variants.

There are still a few issues that need solving but I thought I'd mention it here to get wider feedback. Is it useful, what would make it more useful, etc?   Jts1882 | talk  16:34, 22 February 2019 (UTC)

Pretty neat. So it only shows the taxonomy templates involved, so no species for the most part? --Nessie (talk) 03:36, 23 February 2019 (UTC)

Yes, it only looks at the taxonomy templates, so only those with templates for the species will appear. These seem to be mostly viruses.

It might be possible to add an option to look for species under a genus using a search for speciesbox templates with parameters |genus=GENUS or |taxon=GENUS XXXX). But that goes beyond the current scope. What I want to use it for is to identify inconsistencies in the taxonomies, which are often only partially updated, leaving a hybrid tree.   Jts1882 | talk  08:05, 23 February 2019 (UTC)

I think it's better to keep this very useful tool to its primary purpose, i.e. browsing the taxonomy templates.

Re species taxonomy templates, viruses are a special case, because species don't have binomial names, so the parent taxon can't be derived from the first word (i.e. the genus), as it can for non-viruses. Having said that, some of the species taxonomy templates for viruses can be avoided – see Template:Virusbox/doc. Peter coxhead (talk) 10:08, 23 February 2019 (UTC)

@NessieVL: If you want to find species with a {{speciesbox}} template for a particular taxon you can use the following search:

hastemplate:speciesbox insource: /[gt][ea][nx][uo][sn][ ]*=[ ]*Panthera/

This example for Panthera picks up 17 articles with such taxoboxes, 12 with |genus=Panthera and 5 with |taxon=Panthera. --   Jts1882 | talk  14:39, 26 February 2019 (UTC)

Or even simpler.

The 17 species below Panthera

 hastemplate:Taxonomy/Panthera hastemplate:Speciesbox  

All 61 speciesbox, subspeciesbox and hybridbox using of the genus Panthera taxonomy template

hastemplate:Taxonomy/Panthera  

--   Jts1882 | talk  15:03, 26 February 2019 (UTC)

This is great. There should definitely be a way built into the Template:Taxonomy pages that lets you view their children. It's an active hindrance to development of the automated taxobox system not to have one. Ideally, there should be a full tree of taxonomy templates displayed somewhere. SUM1 (talk) 00:35, 7 April 2019 (UTC)

@SUM1:. I don't think this is possible as we can't use the Wikipedia API for searches from Lua, which means templates can't access this information. What would be possible is to generate the full tree (or several subtrees, as there are >30,000 taxonomy templates) and save them in subpages of the documentation or project pages. This would need periodic updating.   Jts1882 | talk  07:14, 7 April 2019 (UTC)

@SUM1: it's also important to understand that the taxonomy templates do not form a coherent tree, but rather a tangled network or web. By using skip and variant taxonomy templates, the system holds overlapping and inconsistent taxonomies, because this reflects usage in reliable sources – there is absolutely no consensus for a single classification system for the tree of life. The best known variation is probably the use of different classifications for extant birds and dinosaurs, although there are many others. If you root the tree at Chordata, say, you'll find the immediate children are the differently ranked Class Aves, Subphylum Cephalochordata, clade Olfactores, Class Pteraspida and Subphylum Tunicata. By a long route, you can eventually use taxonomy templates to get down to Aves from Olfactores, so the branches join up. (It's not just animals; there are several overlapping systems used for plants, some including Kingdom Plantae, some not.) So a display of the whole system would not be as useful or informative as you might suppose. Peter coxhead (talk) 09:59, 7 April 2019 (UTC)

@Peter coxhead: I'm aware of the nature of overlapping taxonomies and never said there is or should be one single tree. What I meant was it would be extremely useful to be able to view the relationships of all existing Taxonomy templates, for example so that obsolete or synonymous ones can be reorganised and new ones can be added where missing. A taxon might point to a certain parent on its page even though a more immediate parent has already been created, and we wouldn't know about it without some digging. Jts1882's userscript is definitely a big step in the right direction. SUM1 (talk) 14:48, 7 April 2019 (UTC)

@SUM1: You can get the immediate children with my script easily enough. If you use the taxonomy browser from a template page it defaults to that page. You can run it from a parent taxon to see if their are some intermediate levels.

My draft version has an export function. I used it on order Carnivora to 11 levels and have placed the tree on one of my user pages: User:Jts1882/taxonomybrowser/Carnivora (ignore the RHS column). Is there some group that you are interested in where this would be helpful?   Jts1882 | talk  15:21, 7 April 2019 (UTC)

@Jts1882: And I appreciate your userscript and will find it useful in making my edits to taxonomies. At the moment, I'm doing the much-needed update and reorganisation of Heterodonta (which you alluded to on this talk page), and I don't currently need a full visualisation of taxonomy templates in that group for the edits I'm doing, but I may do so in the near future, so I'll let you know. SUM1 (talk) 18:21, 7 April 2019 (UTC)

Just to note here that {{Infraspeciesbox}} has been updated so that, like other automated taxobox templates, it automatically italicizes the page title if it should be italicized, with only a few rare exceptions requiring manual italicization. If you find any errors, please report them at Template talk:Infraspeciesbox#Italic title. Peter coxhead (talk) 21:21, 9 April 2019 (UTC)

Per the previous discussion, I have nominated another set of 25 templates for deletion. See Wikipedia:Templates for discussion/Log/2019 May 2. Comment if you like, especially if you have one or more templates that you want to save. I do not expect the deletion to be controversial. – Jonesey95 (talk) 14:12, 2 May 2019 (UTC)

27 June 2019 update

Mammal subprojects with articles tagged for both mammals and subproject:

Some highlights since the last update (on 1 January). Bats, Bivalves and Birds joined the ranks of projects with 90%+ automatic taxoboxes. Cephalopods, Marine life and Tree of life passed the 50% threshold. Spiders was the #2 project in automatic taxobox usage when I started tracking stats (April 10 2017), but has fallen into 10th place; however, this is the first period where there more than 100 spider articles were converted from manual to automatic taxoboxes. Use of Virusbox has doubled since the start of the year. Plantdrew (talk) 18:35, 27 June 2019 (UTC)

Mammals as a whole are poor compared to the other chordate "classes", but primates and rodents (over half of mammals) are very high. Which are the groups that bring the average down? Is it the fossil groups, or is there a neglected extant group or three? I'm guessing the former.   Jts1882 | talk  19:12, 27 June 2019 (UTC)

Fossils account for the majority of mammals with manual taxoboxes, but extant marsupials have been neglected (as well as a few tiny orders: e.g. Scandentia and Macroscelidea). Here's a link to run the search for manual mammal taxoboxes (PetScan is not working well lately, so you may have to try again later if the link doesn't work). While I'm at it, I'll note that amphibians (especially frogs) are bringing the average down for Amphibians and Reptiles. Plantdrew (talk) 19:41, 27 June 2019 (UTC)

Thanks again for this update, Plantdrew! For those folks wondering about those three remaining manual primates, please see Wikipedia talk:WikiProject Primates#Only four manual taxoboxes left and maybe lend some expertise of you can. I'm surpised about Bats and Dogs as those were 100% a month or so ago, so there's been some backsliding. --Nessie (talk) 23:32, 28 June 2019 (UTC)

I think I've converted all the bat taxoboxes. I found 12.   Jts1882 | talk  08:54, 29 June 2019 (UTC)

How are these lists generated? With Petscan? I'd like to be able to make a further breakdown of mammals. I tried to find the missing cat and dog ones with the search and couldn't find anything (perhaps they have been update, but more likely my search is deficient). I found a number of other carnivores, mainly fossil seals and mustelids, which I updated where possible. There are a number of problematic bear subspecies/population articles that defy the automated system.   Jts1882 | talk  08:52, 30 June 2019 (UTC)

I think the Carnivora are now all converted to automated taxoboxes, except these awkward cases. There were about 100, mostly extinct genera, still using {{taxobox}}. Hopefully I got the search right and there are not another batch to find.   Jts1882 | talk  14:38, 30 June 2019 (UTC)

@Jts1882:, for these stats, I do a PetScan search for a particular WikiProject template on the talkpage and taxoboxes. That does pick up parasites and pathogens that are tagged for the project of the organism they infect. Remaining Cat/Dog articles are all pathogens/parasites. I keep my search methodology consistent for compiling the statistics, but it does get a little misleading when a handful of parasites account for most of the manual taxoboxes under a particular project. I do use PetScan searches based on taxonomy categories when I'm actively converting to automatic taxoboxes, but taxonomy categories can also pick up pathogens (and other irrelevancies) if drilled deep enough. Plantdrew (talk) 16:33, 30 June 2019 (UTC)

Rodents, Birds and Bivalves are also at the point where pathogens account for a good chunk of the manual taxoboxes. Plantdrew (talk) 16:41, 30 June 2019 (UTC)

My first attempts to find the articles with manual taxoboxes cats and dogs also picked up parasites. I think I now have a better search using regex. With birds it picks up 52 articles with manual taxoboxes, which are mostly general common names (eagles, waders) or fossil species with uncertain taxonomy. With rodents, I found 66 articles. Unfortunately I can't compare with Petscan as it is still not working.   Jts1882 | talk  07:21, 1 July 2019 (UTC)

PetScan is sort of working. Whatever the problem is, it fluctuates over a fairly short time scale. It might take 5 or 6 attempts over the course of a minute, but I can usually get a PetScan search to run if I persist. I will get 502 errors and erroneous reports of 0 results in the meantime. It is very annoying. Plantdrew (talk) 17:51, 1 July 2019 (UTC)

Hey, pathogens and parasites need love too! It's probably easier to do them 52 or 66 at a time instead of 1,529, 11,202, or 12,268. Glad I'm not the only one having trouble at PetScan tho. Anyone else having trouble with FIST? Is it all WMFLabs? --Nessie (talk) 19:43, 1 July 2019 (UTC)

I created {{Cleanup taxobox}} for articles with problematic automated taxoboxes. I felt editors could use a way to flag articles that may or may not be listed in one of the bot-maintained categories. It also allows novice editors to request assistance with upgrading to automatic, general formatting, or getting lost with taxonomy templates. Ideally it will also show some of the problems that the bots are missing. --Nessie (talk) 17:21, 2 July 2019 (UTC)

Not sure if this is a bug or a feature, but the {{Speciesbox}} on Pseudo-nitzschia australis is dropping the second part of the genus name from the binomial (i.e. Pseudo australis). The genus displays correctly. The box uses |genus=Pseudo-nitzschia. {{Taxonomy/Pseudo-nitzschia}} uses |link=Pseudo-nitzschia. All I can find in ICN is that hyphens must be removed from epithets and fossil genera. --Nessie (talk) 14:17, 9 September 2019 (UTC)

P.S. I think it's screwing up the title italics as well. --Nessie (talk) 14:23, 9 September 2019 (UTC)

The cause is the use of {{Str letter/trim}} in {{speciesbox}}, which is taking alphabetic characters from the beginning of the string. Removing it corrects this case, but I'm not sure this is still needed to strip disambiguation terms or a left over from before the rewrite with {{speciesbox/getGenus}}. Clearly this is something for Peter coxhead.

I think you are correct about hyphens still being allowed in the generic names of extant species. Another question is what is the abbreviated form of Pseudo-nitzschia australis: is it P. australis or P-n australis?   Jts1882 | talk  15:58, 9 September 2019 (UTC)

Well, as an aged person who learnt Greek, the simpler abbreviation should be Ps. australis, since ψ=ps is a single letter in Greek. Peter coxhead (talk) 21:49, 9 September 2019 (UTC)

::: I think the solution is to remove {{Str letter/trim}} from |binomial= in {{speciesbox}} and to add {{first word|{{{2}}}}} to {{speciesbox/getGenus}} (as done in {{speciesbox/getGenus/sandbox}}). This gives the right result in preview for Pseudo-nitzschia australis and Citronella mucronata (which has |genus=Citronella (genus)).   Jts1882 | talk  17:05, 9 September 2019 (UTC) Strike that, it will interfere with {{Taxon info}}.   Jts1882 | talk  17:13, 9 September 2019 (UTC)

• It's a bug. Hyphens are allowed in names under the ICNafp. For genus names, Art. 60 Note 6 applies: "a non-fossil generic name published with a hyphen can be changed only by conservation". So the code should allow them. The immediate problem is that I can't remember all the stripping that {{Str letter/trim}} is required to do; certainly to remove a parenthesised disambiguator, but possibly also "?". An immediate but ugly solution appears to be to replace {{Str letter/trim|{{Speciesbox/getGenus|{{{taxon|}}}|{{{genus|}}}}}}} by {{replace|{{Str letter/trim|{{replace|{{Speciesbox/getGenus|{{{taxon|}}}|{{{genus|}}}}}|-|Đ}}}}|Đ|-}} which will treat hyphens as a special case and will be safe as "Đ" cannot occur in a genus name. Peter coxhead (talk) 18:38, 9 September 2019 (UTC)

I've implemented this fix; it deals with the incorrect genus name in the binomial but doesn't fix the italicization problem. This can be dealt with by a work-around for now, but the whole issue will need fixing properly. I'm surprised it's never been noticed before. Peter coxhead (talk) 21:16, 9 September 2019 (UTC)

By the way, the Pseudo-nitzschia article has a long list of species with disallowed external links which need to be fixed. Peter coxhead (talk) 21:20, 9 September 2019 (UTC)

Ok, the right solution is to replace {{Str letter/trim}} by a new template {{Speciesbox/trim}} that treats a hyphen as a letter. Doing this in {{Speciesbox/name}} also fixes the italicization issue. See Pseudo-nitzschia australis now. Peter coxhead (talk) 21:42, 9 September 2019 (UTC)

I was trying to think of what could need stripping/trimming apart from the standard parenthetic disambiguation. I ran some searches of for various non-alphabetic characters. The searches picked up two hyphenated genus names, a number of "/?" genus names, one underscore (Atlanta peronii), one "<" from an HTML comment (Giant girdled lizard). All {{speciesbox}}es were handled correctly.

I've fixed the external links in Pseudo-nitzschia although some references aren't complete citations.   Jts1882 | talk  10:00, 10 September 2019 (UTC)

Looks great. I guess with only a handful of taxa involved, it was easy to slip by unnoticed. I almost missed it myself. Glad it (atleast seemed line it) was an easy fix. --Nessie (talk) 13:24, 10 September 2019 (UTC)

@Jts1882: there is another case that should be handled. I've had looking at nothogenera on my to-do list for some time. Speciesbox won't handle genus names beginning with ×. There are very few articles on such taxa, so it wasn't a particularly active item on the list, and there's an argument that odd cases are best left to manual taxoboxes. Altering {{Speciesbox/trim}} to deal with such cases is one step, I now see. Part of the problem is the variation in how nothotaxon names are set out: with or without a space after the hybrid sign, with a thin space, with a non-breaking space, etc. The less powerful regular expressions in Lua (particularly the absence of a whitespace match) makes it a bit fiddly to implement. Peter coxhead (talk) 13:32, 10 September 2019 (UTC)

I see 44 articles starting with × and there is Category:Plant nothogenera ( 43 ) with it's subcats totaling 134 articles, so not very few, but not a lot. --Nessie (talk) 14:34, 10 September 2019 (UTC)

@NessieVL: note that most orchid nothogenera articles are like Brassidium and don't have the page title matching the way the genus name is written in the article (nor is the page title italicized). It will be easier to sort out these pages if automated taxoboxes can properly handle nothogenera. Peter coxhead (talk) 19:41, 10 September 2019 (UTC)

Oh yea, I just wanted to give a range. I know from working through all the animal hybrids that people crammed all sorts of square pegs into round holes. I imagine the asiest thing to do was to omit the × entirely. Plus there have got to be some out there that used the letter "x" and not "×". --Nessie (talk) 22:08, 10 September 2019 (UTC)

Couldn't find any nothogenera in the articles that start with "x ".--Nessie (talk) 22:29, 10 September 2019 (UTC)

Only one doesn't have a space (×Reyllopia conollyana) so that can be standardised. Also it might be easier to handle these with {{hybridbox}}.   Jts1882 | talk  15:01, 10 September 2019 (UTC)

It's not too difficult to accept either × Heucherella (with a normal space) or ×Heucherella (with no space). So if we only allow these two options in taxoboxes, I think the standard {{Automatic taxobox}} and {{Speciesbox}} could be fixed to handle nothogenera, with some work. At present, {{Speciesbox/trim/sandbox}} handles them both:

• {{Speciesbox/trim/sandbox|× Heucherella tiarelloides}} → × Heucherella
• {{Speciesbox/trim/sandbox|×Heucherella tiarelloides}} → ×Heucherella

Peter coxhead (talk) 19:41, 10 September 2019 (UTC)

Update: I have updated {{Speciesbox}} to a version that handles species within nothogenera; see, for example, × Astrolista bicarinata. (There's a small bug remaining: the automatically completed name of the taxobox is not italicized correctly. I'll fix this later.) If you notice any problems this update has created, please revert the latest change to {{Speciesbox}}. Peter coxhead (talk) 11:40, 11 September 2019 (UTC)

 Done automatic italicization of the taxobox name now works. I've also moved ×Reyllopia conollyana to × Reyllopia as it's monotypic. × Astrolista bicarinata should be merged with × Astrolista. Peter coxhead (talk) 17:57, 11 September 2019 (UTC)

There are perhaps a large number of taxonomy templates that use a pipe in the link parameter, like |link=Fly|Diptera. Correct me if i'm wrong, but the two reasons for this of in cases of redirects from monotypic clades and from redirects with possibilities. I can't think of how we could run a search for monotypic clades that are no longer monotypic, but for taxa that are no longer mere possibilities it should be possible, right? Does anyone search for these? I imagine checking for if a non-redirect article exists for the text after the pipe, and maybe flagging it? Is this even an issue? --Nessie (talk) 02:03, 24 September 2019 (UTC)

Does this search or this one help in any way?   Jts1882 | talk  07:24, 24 September 2019 (UTC)

Those just seem to show templates with piped links. I'm looking for situations where first a template was made with the piped link, say to avoid a redirect for the genus page that hasn't been properly written. So the template points to the family article. After the template was made, someone converted the genus redirect into a stub, and didn't change the taxonomy template. So any readers looking at species articles, when they click on the genus in the taxobox, they will miss the new stub and be redirected to the family article instead. How do we look for those instances? Where the taxonomy template says |link=FamilyArticle|Genus and "Genus" is no longer a redirect, but an actual article? --Nessie (talk) 14:03, 24 September 2019 (UTC)

A good question. When |link=LINK_TARGET|LINK_TEXT is present, then LINK_TEXT should be a redirect to LINK_TARGET. I don't think this can be checked by a search, but it could (I think) be detected within the code that processes taxonomy templates, and the template page put in a tracking category, although my memory is that checking whether a page is a redirect to another is an expensive operation, and I'm not sure if this would be justified. Peter coxhead (talk) 17:37, 24 September 2019 (UTC)

{edit conflict}}. I'll put this here in the spirit of throwing out ideas.

Yes, I realise that. I was just getting a list as a first step. In your example, what is wanted is to check if "Genus" has an article that isn't a redirect (if it does, we don't want the pipe any more). So it might be possible to do an API search for articles from a list generated from a search for the name after pipe, which should be the name after the forward slash in the template name ({{Taxonomy/Genus}}). But I don't see how to do this.

Another possibility is to use a category for the list in the API search. If there was a Category:Taxonomy templates using a pipe, then this could be using as the list to search for articles that are not redirects. Although the desired category is Category:Taxa with taxonomy templates using a pipe. Again I can't see how to make the connection.

A category seems the first step in the right direction, although I can't see how to use it properly.   Jts1882 | talk  19:37, 24 September 2019 (UTC)

Could it be compared to members of Category:Redirect-Class articles? --Nessie (talk) 19:51, 24 September 2019 (UTC)

Category:Redirect-Class articles is very underpopulated. Relatively few redirects have a WikiProject banner. Article Alerts has some logic that associates redirects with the WikiProjects of their targets. I would love to be able to search for redirects by project using Article Alert logic (maybe in PetScan?). Plantdrew (talk) 20:07, 24 September 2019 (UTC)

I'm trying to get it more populated/ It also requires each talk page template to use a subpage for quality assessment. I'm working my way through the ToL templates now doing that. --Nessie (talk) 21:24, 24 September 2019 (UTC)

Re expensive operations - the expensive-parser-function limit is 500 per page, so if the highest # of redirects checked is < 5, then I wouldn't see that as disqualifying.   ~ Tom.Reding (talk ⋅dgaf)  19:55, 24 September 2019 (UTC)

I think the way to do this is via javascript and the API.

1. Search for all taxonomy template containing a pipe (e.g. the search above)
2. Strip the TAXON from each taxonomy template page title ({{Taxononomy/TAXON}} to create a list of potential page titles. I believe this should always match |link=TARGET/TAXON in the template, although this can be checked.
3. Use this list to look for pages with another search that returns a list of articles and a list of redirects (e.g. something like this). This would need to be done reiteratively due to limit on number of titles.
4. Look for page titles without a corresponding redirect. These should be the articles that are being bypassed by the pipe in the taxonomy template.

Does this make sense?   Jts1882 | talk  08:33, 25 September 2019 (UTC)

Following up Tom.Reding's comment, I think the answer is to check inside Module:Autotaxobox when processing a taxonomy template, and add the template to a tracking category. Given LINK_TARGET and LINK_TEXT, create a title object corresponding to LINK_TEXT. The redirectTarget field of this object will either return false or a title object for the target of LINK_TEXT if it's a redirect. If a title object is returned, its title field should be LINK_TARGET. (See [1].) If not, the taxonomy template can be added to a tracking category for investigation. I think this should work, but I don't have time to implement and test right now.

The checks can be limited to certain ranks, if this turns out to be required, but as of now, it seem to me that there should always (?normally) be a redirect from LINK_TEXT to LINK_TARGET, regardless of the rank. – see below.

The first step is to write and test a simple Lua function that given two arguments, TARGET and TEXT, returns true if they are equal or if the page with the title TEXT is a redirect to the page with title TARGET, or false in all other cases. Peter coxhead (talk) 13:29, 25 September 2019 (UTC)

No, I was wrong. A common case is that LINK_TEXT is the title of a disambiguation page. An example (which I've actually been working on) is LINK_TARGET = Alsophila (plant), LINK_TEXT = Alsophila. So a category showing cases where LINK_TEXT isn't a redirect to LINK_TARGET will have a lot of false positives. And restricting the rank won't be helpful, I think. Peter coxhead (talk) 16:25, 25 September 2019 (UTC)

I've implemented (at Module:Sandbox/Peter_coxhead/Tests|linkCheck) the simple code that can check whether LINK_TEXT redirects to LINK_TARGET:

• re Template:Taxonomy/Polypodiopsida: {{#invoke:Sandbox/Peter coxhead/Tests|linkCheck|Fern|Polypodiopsida}} → true
• re Template:Taxonomy/Alsophila (plant): {{#invoke:Sandbox/Peter coxhead/Tests|linkCheck|Alsophila (plant)|Alsophila}} → false

This could be added to the relevant Lua code so that taxonomy templates for which this test is false get added to a tracking category. However, as noted above, this category would contain all the cases where the link text is the title of a disambiguation page as well as those where there is a real issue to be fixed. So I'm not sure whether it would be worthwhile. What do others think? Peter coxhead (talk) 08:21, 26 September 2019 (UTC)

It's unfortunate the disamgibuation pages make up such a large fraction of the false cases. It rather undermines the goal of generating a manageable category for manual checking by editors. I note that Alsophila has the category Category:Taxonomy_disambiguation_pages so if there was a way of checking the category then those cases could be filtered out. A crude method with title.getContent() and a search for the string "disambiguation" might be possible (it would pick up categories and text saying the page is a disambiguation page.. I'm not sure such measures would be worth it.   Jts1882 | talk  11:43, 26 September 2019 (UTC)

I agree that it's not worth searching the page content. We could create the tracking category temporarily, then manually check all the entries, and then discontinue the tracking, I suppose. Peter coxhead (talk) 14:19, 26 September 2019 (UTC)

I am not a coder, so this may be a dumb idea, but here goes. The main reason for piping the links in the taxonomy templates is to prevent self-directed links from appearing in the taxobox. That's why the instructions in the {{R taxon with possibilities}} family of templates to not use piped links are abrogated. So would it be less expensive to run this in the template? I realize that somehow the taxobox would have to know if a link redirects to itself, but the gadget that highlights links for me can do it. Something to check if ARTICLE_NAME and PARENT_TAXON_REDIRECT_TARGET match, and if they do, then to deactivate the link in the taxobox. If this could be worked out, then piping links in the taxonomy templates would not be needed, and we wouldn't have to update them every time a new article was written. --Nessie (talk) 14:28, 26 September 2019 (UTC)

Establishing a temporary tracking category, followed by an effort to manually check seems like a reasonable approach. Presumably the category would need to re-established periodically (on a time-scale of years) to find further links that need fixing (which would entail combing through the false positives all over again). Personally, I'd prefer to wait about a year to do this, rather than doing it now. It will take around a year for me to finish converting nearly all plants to automatic taxoboxes, and I expect another year will bring automatic taxobox conversion efforts to near completion for most groups (except perhaps insects/leps and groups where automatic taxoboxes are potentially still controversial (e.g. gastropods)). Plantdrew (talk) 16:27, 26 September 2019 (UTC)

Sounds a good plan to me. Peter coxhead (talk) 21:08, 26 September 2019 (UTC)

When a species is incerta sedis in a family and the working/temporary genus name is placed in quotes, is there a way to reflect this in an automated taxobox? For example: "Centrolene" acanthidiocephalum. Loopy30 (talk) 01:17, 30 October 2019 (UTC)

I'm not sure whether quotes can be made to work, but provisional generic placement can be accomplished with a dedicated taxonomy template and |binomial= in {{Automatic taxobox}}; see existing examples here. Plantdrew (talk) 05:21, 30 October 2019 (UTC)

I think the "?" approach is the one that was meant to be used to reflect uncertain placements. It would be possible to add yet more complexity to the code underlying automated taxoboxes to handle quoted genus names, but I think this would be a mistake – the code is already complex, and the cost-benefit trade-off doesn't persuade me. There's nothing wrong with a few special cases being handled by manual taxoboxes. Peter coxhead (talk) 08:32, 30 October 2019 (UTC)

(Incidentally, the three "?" genera listed here show up the current inconsistencies in how the question marks are handled, i.e. whether parenthesized and/or italicized. I do have fixing it on a to-do list somewhere, but there are so few cases, it's not a priority. Peter coxhead (talk) 08:36, 30 October 2019 (UTC))

What about a special rank? The quoted genus is quite common with fossil species (e.g. "Miacis") so isn't an obscure convention. If the taxonomy template had |rank=genus/incertae sedis or something similar it would be relatively simple to handle separately.   Jts1882 | talk  09:40, 30 October 2019 (UTC)

Thanks to everyone for the replies. I'll just use manual taxoboxes in these cases. 'Cheers, Loopy30 (talk) 12:35, 30 October 2019 (UTC)

Should Speciesboxes support displaying a type species. It would only be relevant for monotypic genera. I do sometimes find manual taxoboxes for monotypic genera that include type species data. Typically, when it is present, the displayed type species is a heterotypic synonym of the sole species. If the type species is different from the only species currently included in a monotypic genus, I would expect it to more frequently be a homotypic synonym (basionym/original combination). But I rarely see a basionym given as the type species (which I attribute to Wikipedia fairly frequently giving a combination for the type species that matches the current genus placement rather than the basionym). Plantdrew (talk) 02:13, 5 November 2019 (UTC)

The ICZN mandates the use of the original combination when citing the type species, and this seems to be followed by the sources I read. Article 10 of the ICNafp is not exactly transparently written, but it says "definite inclusion of the type of a name of a species is effected by citation of, or reference (direct or indirect) to, a validly published species name, whether accepted or synonymized by the author". To me this means that the species name used to indicate the type can be any validly published name, not necessarily the basionym. So for an ICNafp name, it seems to me that the value of a type species parameter would usually just be the target of the Speciesbox. But I can see that there is an issue for ICZN names.

It's only a convention to use {{Speciesbox}} for monospecific genera. The alternative is to use {{Automatic taxobox}} with |type_species= where both the original combination and the accepted name could be given (with some convention to separate them). I think most non-specialists would be confused if they saw a Speciesbox which had "Type species: Alpha beta / Species: G. delta / Binomial: Gamma delta" (which would be possible if the combination Gamma beta were not available). Peter coxhead (talk) 10:59, 5 November 2019 (UTC)

Nelumbo (using subdivision_ref)
Scientific classification
Kingdom:	Plantae
Clade:	Tracheophytes
Clade:	Angiosperms
Clade:	Eudicots
Order:	Proteales
Family:	Nelumbonaceae
Genus:	Nelumbo
Species[1]
Nelumbo lutea (Willd.) Pers. Nelumbo nucifera Gaertn.

Nelumbo (using old method)
Scientific classification
Kingdom:	Plantae
Clade:	Tracheophytes
Clade:	Angiosperms
Clade:	Eudicots
Order:	Proteales
Family:	Nelumbonaceae
Genus:	Nelumbo
Species[1]
Nelumbo lutea (Willd.) Pers. Nelumbo nucifera Gaertn.

Asian elephant
Scientific classification
Domain:	Eukaryota
Kingdom:	Animalia
Phylum:	Chordata
Class:	Mammalia
Order:	Proboscidea
Family:	Elephantidae
Genus:	Elephas
Species:	E. maximus
Binomial name
Elephas maximus Linnaeus, 1758
Subspecies[2]
E. m. maximus E. m. indicus E. m. sumatranus

Should there be a parameter to hold a reference for subordinate taxa? It seems odd that there is |diversity_ref=, but nothing for subdivisions. Plantdrew (talk) 02:05, 5 November 2019 (UTC)

I have been putting this information in |subdivision_rank=, but i think having an explicit reference would encourage more editors, especially inexperienced ones, to add a citation for the subtaxa. I imagine it wouldn’t be hard to implement, but I defer that to those with better knowledge of the templates. --Nessie (talk) 02:59, 5 November 2019 (UTC)

I agree. A seperate parameter does encourage use, where as adding it to the |subdivision_rank= is less obvious, although has the same result. The paraphyletic group box already has it as the same function handles subdivisions, includes, excludes and synonyms.   Jts1882 | talk  06:52, 5 November 2019 (UTC)

This is, indeed, straightforward. I've added it to the sandbox versions, and put a taxobox created in this way here as well as an 'old style' one. The ^[1] in the first taxobox comes from |subdivision_ref=, the second from a ref in |subdivision_ranks=. (Note that the "ref markers" against the three taxon names above are because there's code in {{Taxobox/core/sandbox}} to show refs in the taxonomy templates; we discussed this, but never came to any agreement. Problems with the third approach, the ^[2] and ^[3], show up – there's no way of ensuring style consistency, nor that the ref won't be repeated.) I'll try a few more tests, but I see no reason not to make this live soon (minus the "taxonomy template refs" bits). Peter coxhead (talk) 11:36, 5 November 2019 (UTC)

Question: if you put refs in both places (i.e. in |subdivision_ref= and in |subdivision_ranks=), they will all appear. I assume this is what should happen, but I suppose the former could suppress the latter. Views? Peter coxhead (talk) 11:56, 5 November 2019 (UTC)

No need, as an editor adding a reference should notice the duplication. It might be more confusing if the reference on |subdivision_ranks= disappeared. You wouldn't want to suppress a different reference so you'd only suppress them if exactly the same, which is unlikely unless the same editor added them.   Jts1882 | talk  12:35, 5 November 2019 (UTC)

 Done Now live for {{Automatic taxobox}} and {{Speciesbox}} – see third taxobox. Peter coxhead (talk) 16:13, 7 November 2019 (UTC)

@Peter coxhead: Can this also be implemented in {{Virusbox}}? That's probably the only other Autobox that could use it. --Nessie (talk) 16:29, 7 November 2019 (UTC)

I've also made the change to the manual {{Taxobox}}, by the way. I'll look at {{Virusbox}} now. Peter coxhead (talk) 16:34, 7 November 2019 (UTC)

@NessieVL:  Done See the taxobox at Nidovirales (may not be the ref you wanted; fix if not). Peter coxhead (talk) 16:54, 7 November 2019 (UTC)

Nelumbo
Scientific classification
Kingdom:	Plantae
Clade:	Tracheophytes
Clade:	Angiosperms
Clade:	Eudicots
Order:	Proteales
Family:	Nelumbonaceae
Genus:	Nelumbo

Note: the comments below relate to "ref markers" in taxoboxes to mark the presence of references in the taxonomy templates, as per the taxobox shown here. There's code to do this in three ways created by Jts1882 in Module:Autotaxobox/sandbox, accessed via {{Automatic taxobox/sandbox}}.

A belated response on the "ref markers" in the taxonomy listing. Adding extra elements may be confusing for the general reader, but of more interest to those more interested in taxonomy and those checking the automatic taxonomy system. One option is to make the feature optional (opt in). Add these elements (e.g. the little info symbol with the link) within a span with a CSS class making them hidden by default and visible when editors add appropriate code to their common.css. So use <span class="plainlinks taxonomy-template-info" > ... </span> with CCS code .taxonomy-template-info { display:none; }. Then the custom.css would contain .taxonomy-template-info { display:inline !important; }. Or something along those lines.   Jts1882 | talk  07:40, 9 November 2019 (UTC)

If that were the path, I would want the display option button on the box itself. A better option could be to allow the parent refs display to be turned off within the template with something like |display_parent_refs= or something. This would bandage the mixed-reference-style problem. --Nessie (talk) 15:18, 9 November 2019 (UTC)

I think this needs a much wider discussion, as it would affect taxoboxes across all the ToL WikiProjects. Personally, if we do go down the route of displaying markers, I favour something that acts like the middle one: an "info" link to the taxonomy template, but with a smaller icon. Peter coxhead (talk) 17:39, 9 November 2019 (UTC)

Up the now, the system didn't handle names like "Prostherapis" dunni very well. In particular, if you used the obvious {{Speciesbox}}, the genus name wasn't abbreviated correctly. Following a discussion at User talk:Loopy30#"Prostherapis" dunni I've now updated the system to handle such cases (ideally the double quote marks shouldn't be italicized, but I haven't fixed that yet). So there are now two ways of handling a species whose genus name is quoted.

"Prostherapis" dunni
Scientific classification
Missing taxonomy template ():	"Prostherapis" dunni
Species:	[[Template:Taxonomy/"Prostherapis" dunni]]'
Binomial name
[[Template:Taxonomy/"Prostherapis" dunni]]

1) Use {{Speciesbox}} in the normal way

{{Speciesbox
|taxon = "Prostherapis" dunni
}}

or

{{Speciesbox
|genus = "Prostherapis"
|species = dunni
}}

This will link to the genus taxonomy template, as normal, i.e. Template:Taxonomy/"Prostherapis" in this case. There you can set whatever parent is appropriate.

"Prostherapis" dunni
Scientific classification
Missing taxonomy template ():	"Prostherapis" dunni

2) You may not want to display the genus in the taxobox, since it's really incertae sedis. To achieve this, use {{Automatic taxobox}} in the form

{{Automatic taxobox
|taxon = "Prostherapis" dunni
}}

This will link to the species taxonomy template, i.e. Template:Taxonomy/"Prostherapis" dunni in this case. There you can set the parent to be the rank above the genus. (Note that the species name isn't abbreviated in this taxobox; it's only appropriate when the line above shows the genus.) (Or you could set the parent to a genus displayed as incertae sedis.)

I've posted here before updating the documentation in case there are any comments. Peter coxhead (talk) 17:06, 11 November 2019 (UTC)

Is an abbreviation like "P." dunni standard, or an artifact? --Nessie (talk) 20:08, 11 November 2019 (UTC)

@NessieVL: good question. I think it looks odd to repeat "Prostherapis" three times, which is what would happen otherwise with a Speciesbox (which didn't work at all before). I would tend to use the second style, I think, but editors now have a choice. Peter coxhead (talk) 20:25, 11 November 2019 (UTC)

@Peter coxhead: Fair enough. The other difference I see is that in the latter "Prostherapis" is never linked. That can be worked around by linking in the prose, but something to be aware of. Could this approach also be applied to other oddball orthographies like the ones in Category:Automatic taxoboxes using binomial parameter ( 0 )? (btw we need to update the Binomial nomenclature article to explain all this punctuation). --Nessie (talk) 14:50, 12 November 2019 (UTC)

@NessieVL: the second style only makes sense if there's one species in the quoted genus, I think, so there wouldn't be a link anyway.

I made this change in response to comments made at User talk:Loopy30#"Prostherapis" dunni. I wouldn't have implemented it if it had involved too many changes to the existing code, but it turned out that basically skipping the " characters was pretty straightforward.

Some of the 'oddball' articles in Category:Automatic taxoboxes using binomial parameter are "Candidatus" articles. As can be seen from Category:Candidatus taxa ( 123 ), there are enough articles of this kind to justify having a way to handle "Candidatus" genera and species. However, how to present such taxa in taxoboxes isn't clear; several different styles are currently being used in manual taxoboxes. It should really be discussed at Wikipedia talk:WikiProject Microbiology, but this seems to be an inactive project. Peter coxhead (talk) 15:41, 12 November 2019 (UTC)

I spent some time on those taxa and set up that category to at least keep track of them. The main issue for displaying them is that the Latin (Candidatus or Ca.) is italicised but the genus and epithet are NOT. So they should be written like "Candidatus Liberibacter solanacearum" (with quotes). --Nessie (talk) 16:09, 12 November 2019 (UTC)

@NessieVL: the category (which I've only recently found) is very useful. The styles do vary widely, though. For example:

• Candidatus Legionella jeonii uses a manual taxobox, and currently has the article title as all italics (which can be fixed by changing {{Italic title}} to DISPLAYTITLE:). It does not have quotes either in the article or the taxobox.
• Candidatus Phytoplasma fraxini uses Speciesbox, and currently has the article title as all nonitalics. It does not have quotes anywhere, and does not display the binomial with "Candidatus" or "Ca." because automated taxoboxes don't handle such names.
• Candidatus Caballeronia kirkii uses a manual taxobox and is one of the few articles I've found that are formatted as you say it should be.
• Sulcia muelleri uses Speciesbox, and doesn't have "candidatus" in the article title or the taxobox. It shows up in the error-tracking category Category:Taxonomy templates with name and link text not matching because the taxonomy template does have "Candidatus" in its title, but this doesn't match the article.

Basically the articles are a mess and a pain! I favour using manual taxoboxes throughout for now. I think that if an automated taxobox is really wanted, it will need to be a new one ("Candidatusbox"?). The same argument applies as for viruses, I believe: coping with a different style of italicization (plus in this case a prefix) would make adapting the underlying code of the main automated taxoboxes too complex. Peter coxhead (talk) 16:56, 12 November 2019 (UTC)

@Peter coxhead: hence why the can got kicked down the road. . --Nessie (talk) 17:13, 12 November 2019 (UTC)

•has Nessie asked why the quotation marks would be italicised? ~ cygnis insignis 17:07, 12 November 2019 (UTC)

@Cygnis insignis: Should question marks be italicized? --Nessie (talk) 17:13, 12 November 2019 (UTC)

never, in fact, never-ever ~ cygnis insignis 19:52, 12 November 2019 (UTC)

How about instead of trying to make automatic taxoboxes format various nonstandard and rare use cases correctly (or adding yet another taxobox template for Candidatus), adding a parameter (or two) that allows manual formatting and overrides the automatic format? Edible frog has a manual taxobox with binomial=Pelophylax kl. esculentus and species=P. lessonae × P. ridibundus; add |binomial_formatted= and |species_formatted= which would take the values currently given in edible frog and display them as they appear now. Plantdrew (talk) 17:35, 12 November 2019 (UTC)

@Plantdrew: That sounds a lot better than creating {{Kleptonbox}}. How hard would that be to implement? --Nessie (talk) 19:00, 12 November 2019 (UTC)

@Plantdrew and NessieVL: yes, a very good idea, I think. To handle "candidatus" would also need |genus_formatted=. So the idea would be to use |taxon= in {{Automatic taxobox}} to enter the automated system at the level of genus or above, and then supply the missing ranks below this level manually. You would usually also have to format the article title manually and supply a formatted |name=, I believe. I don't think it should be too difficult to implement – at least without any error-checking for duplicated or missing ranks. It's certainly worth trying out. Peter coxhead (talk) 19:17, 12 November 2019 (UTC)

I have another solution, if the display can be modified for species names ~ cygnis insignis|

It turns out (with some testing) that it will be easy to add "manual" values for the genus, species and/or binomial, picking up the higher ranks from |taxon=. The tricky bit to work out, I've discovered, is matching up the authorities. Automated taxoboxes use position from the target taxon upwards (via |authority=, |parent_authority=, |grandparent_authority=, etc.).

It's not clear to me what to regard as the target taxon and hence the rank to which |authority= is attached when there are "manual" values: the lowest of the automated and manual ranks? the automated rank? Suppose the automated rank is the genus and we are supplying a formatted species name. Should the taxobox have something like

|taxon = [a genus]
|authority = [for the genus]
|species_formatted =
|species_formatted_authority =

or something like

|taxon = [a genus]
|parent_authority = [for the genus]
|species_formatted =
|authority = [for the species as it's the lowest rank, i.e. the target taxon]

Both can be coded, although both are fiddly, because somehow all authorities have to be connected to their ranks before being displayed. Peter coxhead (talk) 09:55, 13 November 2019 (UTC)

The second example, "|parent_authority = [for the genus]", please. I'm having trouble envisioning where the genus authority would be worth displaying when manual formatting of the species is useful. Are there monotypic Candidatus "genera"? Genus authority isn't useful in Edible frog or "Prostherapis" species, is it? There may be relevant authorities for Candidatus species, species in quoted genera, in which case "|authority = [for the species as it's the lowest rank, i.e. the target taxon]" would be most appropriate. Manual formatting (to bypass some italics) could also be useful for applying automatic taxoboxes to articles in Category:Undescribed species which shouldn't have an authority at all.Plantdrew (talk) 04:34, 14 November 2019 (UTC)

As a general point on authorities, they are vulnerable to changes in the taxonomy templates. If someone sets up parent, grandparent and greatgrandparent authorities on a genus to refer to family, order and class, the assignment would be changed if another editor changed the taxonomy templates to introduce another level, say subfamily, or if the family assignment was changed in the taxonomy template. Such changes would be transparent to some one watching the page with the taxobox. Although I can't think of a solution, short of using the taxonomy template for authorities, it might be worth thinking about if changes are to be made to authority handling code.   Jts1882 | talk  07:35, 14 November 2019 (UTC)

@Plantdrew: yes, on further reflection, to maintain consistency with the way that all the other automated taxoboxes work, the convention should be maintained that the lowest rank has |authority=, the next up |parent_authority=, etc.

@Jts1882: yes, the convention for matching authorities with taxa for monotypic taxa is problematic, as you say. However, there doesn't seem to be a better method:

• Switching to rank-based parameters doesn't work because ranks like "clade" occur many times, especially in deep hierarchies. Also ranks are equally liable to be changed in the taxonomy template once you get above genus and family; superfamilies get treated as clades, etc.
• I've been attracted to putting the authorities in the taxonomy templates in the past, but actually this solves nothing because you would still need to put something in the taxobox to indicate that the authority should be displayed if taxoboxes are to conform to the status quo (higher rank authorities only displayed for monotypic taxa), and this indication would be subject to exactly the same problem of how to specify which rank it applied to. (There are also issues about what is meant by an "authority" for ranks not covered in the nomenclature codes.)
• If we were starting again, we could put an authority in the taxonomy template and always display it, but the work required to switch over all the taxonomy templates and the taxoboxes in articles is unacceptable, even if there were agreement to do it (and I personally prefer the status quo).

So I think we are stuck with the present system, as the least worst. Peter coxhead (talk) 10:03, 14 November 2019 (UTC)

Ok, so here is a sandbox version that works. For now, there are two "manual" ranks allowed: |genus_text= and |species_text=. These are just the parameter names I chose for now; we can discuss whether there are better ones. There should be a bit more error-checking (e.g. if you supply |species_text=, you need to supply |binomial=). Examples:

Edible frog
Scientific classification
Domain:	Eukaryota
Kingdom:	Animalia
Phylum:	Chordata
Class:	Amphibia
Order:	Anura
Family:	Ranidae
Genus:	Pelophylax
Species:	P. lessonae × P. ridibundus
Binomial name
Pelophylax kl. esculentus (Linnaeus, 1758)
Synonyms
Pelophylax esculentus (Linnaeus, 1758) Rana esculenta Linnaeus, 1758

{{Automatic taxobox/sandbox
| name = Edible frog
| taxon = Pelophylax
| species_text = ''[[Pool frog|P. lessonae]]'' × ''[[Marsh frog|P. ridibundus]]''
| binomial = ''Pelophylax'' [[klepton|kl.]] ''esculentus''
| authority = ([[Carl Linnaeus|Linnaeus]], [[10th edition of Systema Naturae|1758]])
| synonyms =
''Pelophylax esculentus'' <small>([[Carl Linnaeus|Linnaeus]], [[10th edition of Systema Naturae|1758]])</small><br />
''Rana esculenta'' <small>[[Carl Linnaeus|Linnaeus]], [[10th edition of Systema Naturae|1758]]</small>
}}

"Candidatus Brocadia anammoxidans"
cell diagram of Brocadia anammoxidans
Scientific classification
Domain:	Bacteria
Phylum:	Planctomycetota
Class:	Planctomycetia
Order:	Planctomycetales
Family:	Planctomycetaceae
Genus:	"Ca. Brocadia"
Species:	"Ca. B. anammoxidans"
Binomial name
"Candidatus Brocadia anammoxidans"

{{Automatic taxobox/sandbox
| name = "''Candidatus'' Brocadia anammoxidans"
| image = Brocadia anammoxidans.jpg
| image_alt = cell diagram of "Brocadia anammoxidans"
| image_caption = cell diagram of ''Brocadia anammoxidans''
| taxon = Planctomycetaceae
| genus_text = "''Ca.'' Brocadia"
| species_text = "''Ca.'' B. anammoxidans"
| binomial = "''Candidatus'' Brocadia anammoxidans"
}}

Peter coxhead (talk) 17:59, 14 November 2019 (UTC)

@Peter coxhead: Great work as always on a task that was just dumped on you. I think something like |species_display= instead of |species_manual= might be clearer. Looks pretty accommodating though. One still needs to use DISPLAYTITLE, or is just |name= ok? --Nessie (talk) 19:38, 15 November 2019 (UTC)

@NessieVL: maybe "_displayed" is better? It fits more with Plantdrew's suggestion of "_formatted", which I thought didn't quite capture the idea that this was a manual addition, not just a formatting of an automatically handled rank. (It's a pity we can't use the plain |genus= and |species= but these are used for the automated rank in {{Speciesbox}}, so would be confusing – also it's a common error when converting a manual taxobox to leave one or other of these in place of |taxon=, and we need to be able to catch such slips.)

I'll wait and see if there are any more comments before making it live. We should then be able to convert all the "candidatus" cases. Peter coxhead (talk) 09:55, 16 November 2019 (UTC)

Maybe |genus_text= and |species_text= would describe the parameters better. Should the binomial also have its own parameter to make clear that it is overriding the automatic handling?

The only other comment is it does seem a bit counterintuitive to have this handled by {{automatic taxobox}} rather than {{speciesbox}}. It's a species and |taxon= no longer refers to the taxon. On the other hand, these exceptional cases are rarely that intuitive and need to rely on some compromise.   Jts1882 | talk  10:50, 16 November 2019 (UTC)

@Jts1882: I would definitely have binomial_... as an alias (as it happens |binomial= can already be used to override the automatic handling).

I understand the point about {{Speciesbox}}, but (a) this approach is also needed for the special handling of genera, and it seems better to keep the required code together (b) longer term, I think {{Speciesbox}} should become a redirect, and {{Automatic taxobox}} should handle taxa from species upwards (it's probably better to keep {{Subspeciesbox}} and {{Infraspeciesbox}} to mark the different way in which these ranks are handled in the two main codes). When Smith609 created {{Speciesbox}} in 2011‎, the logic and text processing needed to handle uninomial and binomial names within one template would have been tedious and inefficient to code in the template language. In Lua (as you know), it's much easier. Peter coxhead (talk) 15:05, 16 November 2019 (UTC)

Oh, and note that {{Automatic taxobox}} is already used for special cases like Deinodryinus? aptianus, which can in future be handled by the new approach and the italization of the "?" fixed. Peter coxhead (talk) 15:10, 16 November 2019 (UTC)

|genus_text= sounds ok too. I assume there will be tracking categories? --Nessie (talk) 13:29, 16 November 2019 (UTC)

@NessieVL: definitely; just as we have Category:Automatic taxoboxes using binomial parameter now – indeed these cases can be included in a new broader category, something like "Category:Automatic taxoboxes using manual parameters".

@Jts1882 and NessieVL: I'm all in favour of less typing, so _text appeals to me, if others are happy with it. Peter coxhead (talk) 15:05, 16 November 2019 (UTC)

Quassia sp. 'Moonee Creek'
Scientific classification
Kingdom:	Plantae
Clade:	Tracheophytes
Clade:	Angiosperms
Clade:	Eudicots
Clade:	Rosids
Order:	Sapindales
Family:	Simaroubaceae
Genus:	Quassia
Species:	Q. sp. 'Moonee Creek'
Binomial name
Quassia sp. 'Moonee Creek'

Looks good to me. I'm fine with _text parameters (and don't think my "formatted" suggestion for the name was very good). Plantdrew (talk) 02:46, 18 November 2019 (UTC)

• Something we need to keep in mind, once this change goes live and we get through the pile of taxa that need to use this: this opens the door for all sorts of wackiness. We all need to watch the maintenance category because vandalism and misteaks will be easy to overlook in these new parameters. --Nessie (talk) 03:57, 18 November 2019 (UTC)

@NessieVL: absolutely. Peter coxhead (talk) 14:48, 18 November 2019 (UTC)

 Done Now live; no problems with 'ordinary' taxoboxes seen so far. See Category:Automatic taxoboxes using manual parameters for some articles that I have converted to use the new approach. Documentation to follow. Peter coxhead (talk) 15:47, 18 November 2019 (UTC)

I found a small problem, now corrected, with the lowest automatically handled taxon being in bold when it shouldn't have been. Otherwise, I've edited the articles in Category:Automatic taxoboxes using manual parameters purely as a test; I don't intend, for example, to edit all the "candidatus" taxon articles to use the new approach. Peter coxhead (talk) 17:24, 18 November 2019 (UTC)

I've updated the documentation to describe the new parameters at Template:Automatic taxobox#Exceptional cases. Please have a look and amend as needed. Peter coxhead (talk) 11:37, 19 November 2019 (UTC)

Can the list of options for the "status_system" be listed here (Template:Automatic taxobox) and Template:Speciesbox, etc? Wikipedia:Conservation status? —Hyperik ⌜talk⌟ 20:44, 9 January 2020 (UTC)

The documentation for {{Taxobox/species}} has an incomplete listing of options.   Jts1882 | talk  14:26, 27 January 2020 (UTC)

Mammal subprojects with articles tagged for both mammals and subproject:

The big news of the last 6 months is that automatic taxoboxes are now used on more articles than manual taxoboxes. Plantdrew (talk) 21:26, 1 January 2020 (UTC)

Bump (mentioned at Wikipedia:WikiProject Tree of Life/Newsletter/010, adding this comment to prevent archiving until next edition of newsletter is published). Plantdrew (talk) 21:13, 4 February 2020 (UTC)

How can I find these articles listed? I can find 13 articles with manual taxoboxes and a mention of Felidae in the text, but none of these are about felids (they are mostly parasites). Similarly for dogs and 9 articles. The ones I checked don't seem to have relevant projects or categories.   Jts1882 | talk  08:15, 5 February 2020 (UTC)

@Jts1882: see Wikipedia talk:Automated taxobox system/Archive 1#Automatic taxobox usage by project and following threads with "statistics" in them in the archives. WikiProject tags get added to a wide range of supposedly relevant articles. Peter coxhead (talk) 09:00, 5 February 2020 (UTC)

Counts are based on PetScan searches for WikiProject banners and taxobox templates. Pathogens and parasites may have banners related to the organisms that they infect. For projects with >95% automatic taxoboxes, remaining manual taxoboxes are almost all on articles about parasites/pathogens, common names (paraphyletic groups), or fossil taxa. I perhaps should have excluded parasites from the counts, but a) that would be complicated and b) I want to keep the method for compiling the statistics consistent with how I've been doing it since 2017. Plantdrew (talk) 17:34, 5 February 2020 (UTC)

I've found the petscan searches (when they work). I think you are right that separating included taxa and other articles is too difficult and consistency of the analysis is more important. I've converted the cat parasites to automated taxoboxes and intend to do the same for dogs (when Petscan works).   Jts1882 | talk  20:17, 5 February 2020 (UTC)